Hi everyone, I wanted to offer a fresh perspective on the Pocket Reform Next for companies, so that hopefully more companies feel confident to take this route.
I run a software company, we are almost entirely a developer shop. Everyone works remotely, and we get together in person only a couple of times per year. That shape of company drives a very specific hardware profile:
* **ARM everywhere.** Our entire laptop fleet is ARM based. Until now that meant Snapdragon X machines. We are not nostalgic about x86 and we are not going back.
* **Small footprint.** People travel constantly, between client sites, conferences, and the two yearly offsites. A 13 inch laptop is already too big for some of the team.
* **Real keyboards.** A lot of the devs type in Colemak or Dvorak on low profile mechanical Nuphy boards. The internal keyboard of any laptop is, at best, a fallback, but swappable keycaps are a must-have.
* **Heavy lifting lives in Azure.** For builds, model work, and anything GPU bound, every employee gets an Azure Virtual Desktop session on the NGads V620 series (AMD Radeon PRO V620). The local device does not need to be a workstation. It needs to be a great thin client and a great dev box for editor, terminal, and browser work.
* **Display freedom.** Most devs run either an external monitor at their home desk or XR glasses from XREAL or Viture on the road.
When you take that whole profile and try to move it to Linux on ARM, you very quickly discover something uncomfortable: there is almost nothing on the market. Snapdragon Linux support is still a moving target, the Apple Silicon route is a non starter for a managed corporate fleet, and the usual SBC in a case experiments do not survive contact with a travelling consultant.
The MNT Reform family, and in particular the **Pocket Reform Next** with the RK3588 module, is genuinely the only serious answer we found.
Not “one of the best”. The only one.
Small, ARM native, fully documented, repairable down to the screw, and shipping with a software stack that does not fight you when you want to replace pieces of it. For a company that plans to keep devices in service for years and wants to be able to fix them in house, that combination is in a category of its own.
So we committed, and we built a golden image.
## What our Pocket Reform Next fleet image looks like
The goal was not “Linux on a cute ARM laptop”. The goal was a proper enterprise endpoint that happens to be a Pocket Reform Next. After a few weeks of iteration on a single unit, we now have a reproducible image that gives us:
* **Centralized management through an Intune equivalent built on Azure Arc.** The device onboards as an Arc enabled server, reports inventory and compliance, and receives configuration and policy through the same Azure control plane we use for the rest of our infrastructure.
* **Entra ID login at the display manager,** including FIDO2 security keys. Users sign in with their corporate identity, not with a local account.
* **Entra backed sudo.** Local administrator rights are gated by membership in an Entra group, so the same governance model we use everywhere else applies here too.
* **A custom SDDM greeter** themed for the company.
* **A custom Hyprland session** as the default desktop, with our own waybar, swaync, rofi, wallust, kitty, wallpapers, and scripts, all delivered through a single skel snapshot so a freshly provisioned user gets the full experience on first login with zero manual steps.
* **Managed Chromium** for the browser, with company PWAs (Microsoft 365, Teams, our internal apps, the AVD HTML5 client) synchronized centrally.
* **LUKS root with an initramfs key flow** that keeps the unlock experience sane while still meeting our at rest requirements.
The end result is, for us, our own little OS. It is recognisably the company’s environment from the moment the lid opens, and it is built entirely on top of the upstream MNT stack without forking it.
## Two open items before we kick off the rollout
We are honest about the fact that the project is not 100 percent green yet. Two items are still open, and both are well understood:
1. **Hardware accelerated video decode through the Mali GPU and the RK3588 video pipeline.** This matters for us not because of YouTube, but because of **Microsoft Teams** running as a Chromium PWA, the AVD HTML5 client, and similar daily tools. We have a working v4l2 request path for native players, but the Chromium side still needs the rkvdec2 plus LibYUV patch set landed in a packaged Chromium build before we can call it done.
2. **USB C DisplayPort Alt Mode on one of the USB C ports.** Today we can drive an external display either through the Micro HDMI port or through a DisplayLink adapter on USB C, but both paths require an external power source, which defeats the “one cable to the desk, one cable to the glasses” story we want for XR and for hot desking. On the current kernel and DTB the RK3588 DP controller nodes are still disabled and the Type C policy plumbing is not exposed, so Alt Mode simply is not there yet.
The very good news is that **Lucie already has a working solution** for the DP Alt Mode side. We are eagerly waiting to pick it up the moment it is released and fold it into our image. That single change unlocks the XR glasses use case for the whole company.
## Why this matters beyond our company
The Pocket Reform Next is usually framed, fairly, as a device for individual nerds, hackers, and tinkerers. What we wanted to share with the community is that with a reasonable amount of integration work, the exact same machine becomes, in our view, **the perfect enterprise computing device for 2026**: ARM native, genuinely portable, fully owned by the company that deploys it, repairable on a kitchen table, and free from the planned obsolescence that defines most of the corporate laptop market.
We did not have to compromise on identity, management, security, or developer experience to get there. We just had to do the work, and the platform let us.
Thank you to the MNT team and to everyone in this community whose notes, patches, and dts experiments we have been reading for months. We will publish our setup once it is stable enough to be useful to others, and we will of course feed back anything we learn along the way.
Looking forward to Lucie’s DP Alt Mode work, and to running Teams calls on a fully decoded Mali pipeline. Until then, the device is already, by a wide margin, the best laptop any of us has owned.